Reframing crops as multifunctional: a trait-based approach for sustainable agriculture

Global demand for food continues to rise, yet the ecological foundations sustaining crop production, including soil health, nutrient cycling, trophic interactions and biodiversity, are increasingly degraded^1,2. Agricultural performance remains largely defined by yield, a narrow focus that has contributed to ecological simplification, genetic erosion and dependence on synthetic inputs^3,4. Growing awareness of ecological trade-offs has highlighted the role of biodiversity and ecosystem services in supporting crop production^5,6. This recognition has driven the development of sustainable approaches such as ecological intensification, which explicitly leverage ecosystem contributions to reduce yield gaps and dependence on external inputs^7,8.

Within this framework, however, crops are predominantly conceptualised as recipients, or sinks, of ecosystem services: organisms that depend on pollination, nutrient cycling, soil biological activity, pest suppression and microclimatic regulation generated by associated biodiversity and landscape structure^7,9. Yet crops are also active contributors to regulating, supporting and cultural services (Fig. 1), a dimension that remains largely understudied. Crop contributions vary among species and varieties according to their growth forms, physiological strategies and life cycles, with consequences for carbon and nutrient dynamics, microclimate, biodiversity and cultural values^10,11. Recognising the dual role of crops as both sinks and sources of ecosystem services is crucial for understanding how species and varieties shape agroecosystem functioning and provides the conceptual basis for assessing crop multifunctionality¹².

Attempts to evaluate crops from a multiple ecosystem service perspective have been made, particularly in multispecies systems such as cover crop mixtures, polycultures, intercropping and agroforestry^13,14,15. However, the contribution of individual crop species or varieties remains difficult to isolate, and analyses often focus on limited sets of services^16,17. Assessing how individual crop species and varieties contribute to ecosystem services is increasingly important. Cultivated plants harbour substantial functional and genetic diversity, yet much of their ecological potential remains underexplored beyond yield-focused goals^18,19. Breeding alters traits that influence ecological functioning, but the ecological consequences of such breeding decisions are seldom quantified²⁰. In addition, the ongoing homogenisation of global crop germplasm threatens the functional diversity that underpins resilience, pest regulation, soil functions and cultural values²¹. Making the full range of crop contributions explicit and comparable, from yield and biodiversity support to soil functioning, microclimate regulation and cultural values, is therefore critical for identifying overlooked but functionally important species and varieties, and for informing breeding, management and agroecosystem design. With this Perspective paper, we address this gap by proposing a trait–process–service framework that links functional traits to ecological processes and service delivery, enabling systematic and measurable comparisons of crop multifunctionality across species and varieties.

Crops express a wide range of morphological, physiological and biochemical traits that shape both their responses to environmental and management conditions and their contributions to ecological processes²². In functional ecology, this distinction is captured by two main categories of traits. Response traits determine how plants cope with abiotic and biotic stress, resource availability and disturbances. Effect traits, in turn, govern how plants influence ecosystem functioning through their impacts on ecological processes^10,11,23. For example, response traits related to phenology, stress-tolerance mechanisms, biomass-allocation patterns and inducible defences shape how crops respond to environmental conditions. These traits regulate crop responses to climatic and edaphic constraints as well as susceptibility to pathogens and herbivores²⁴. By contrast, effect traits related to residue quality, plant architecture and floral characteristics shape processes such as carbon cycling, microclimate regulation, soil structure and species interactions^25,26,27,28. Importantly, some traits simultaneously act as response and effect traits (e.g., canopy and root structure), influencing crop performance while also shaping ecosystem processes^29,30. Despite its relevance, the response–effect trait distinction remains weakly integrated into agricultural research and breeding, which has largely prioritised yield potential and stability, stress tolerance and resource-use efficiency over traits regulating other ecological processes³¹. As a result, breeding remains only partially aligned with the multifunctional ecological demands of farmers’ fields^20,32.

Building on the distinction between response and effect traits, we focus here on how crop effect traits translate into ecosystem services through specific ecological processes^10,11,33. This provides a structured basis for linking crop characteristics to service provision across species and varieties. An illustrative (non-exhaustive) synthesis of trait–service relationships is provided in Table 1.

Genetic resources

Genetic diversity underpins variation in morphological, physiological, phenological, reproductive and defence traits, providing crops with the response capacities needed to persist across environments and the ability to influence ecological processes. Differences among species and varieties expand the pool of available functional traits, strengthening adaptability and multifunctionality³⁴. The use of landraces, defined as genetically diverse locally adapted crop populations, provides access to broader gene pools than modern commercial varieties, encompassing a wider range of functional traits. For example, bread wheat landraces can show greater root biomass and deeper root distribution than modern cultivars, retaining functional variation relevant to belowground resource acquisition and soil processes³⁵. Genetic resources thus support not only classical breeding for improved crop productivity³⁶, but also the design of crop species and varieties with high multifunctionality or complementary ecological functions.

Climate regulation

Climate-regulating services emerge from traits that modify microclimate, hydrological dynamics and carbon fluxes. Canopy structure, leaf morphology, rooting depth and architecture and litter quality jointly regulate evapotranspiration, shading, infiltration and water redistribution within the soil profile, shaping surface temperature and moisture regimes^37,38. These traits also determine the quantity and persistence of carbon stored both below and above ground^10,39. For example, deep-rooted crops enhance carbon sequestration through greater root turnover and rhizodeposition, while traits associated with tall, perennial or woody growth forms support aboveground carbon storage through the accumulation of persistent biomass⁴⁰. Several of these traits also mediate crop responses to heat and drought, reinforcing local microclimatic and hydrological regulation⁴¹.

Pollination, pest control and aboveground biodiversity

Aboveground services depend strongly on crop effect traits that create resources and habitat structure for pollinators, herbivores and natural enemies. Floral morphology and display size, nectar production and volatile emissions determine the identity and diversity of pollinators that crops support. For instance, corolla depth and nectar accessibility filter pollinator identity and body size, shaping crop visitation and network structure⁴². Structural and chemical leaf traits, including toughness, trichomes and secondary metabolites, shape herbivore community composition and potentially the balance between generalist and specialist feeders, with cascading bottom-up effects on predator and parasitoid communities⁴³. Architectural features such as branching patterns and floral and extra-floral resources provide shelter and trophic subsidies for natural enemies, facilitating biological control⁴⁴. Species with greater floral area and accessible composite flowers, for instance, consistently attract more parasitoids and predators^43,45. The same functional attributes also mediate crop vulnerability to pests and pollinator dependence as response traits while actively shaping the structure and diversity of aboveground biotic communities as effect traits.

Soil formation, nutrient cycling and belowground biodiversity

Soil services arise from plant traits that regulate carbon inputs, soil structure, organic matter dynamics and rhizosphere interactions. Root architecture, exudation profiles and turnover shape soil aggregation, pore development, microbial communities and nutrient cycling^10,38,46. Continuous deposition of fine roots, rhizodeposits and above- and belowground residues fuel microbial activity, influencing decomposition dynamics^25,28. Deep or coarse root systems act as bio-ploughing agents, creating biopores and improving soil structure, while differences in carbon inputs and residue quality modulate sequestration, decomposition and organic matter stabilisation^37,40. Research on cover crops illustrates how combining deep, coarse-rooted species with fine-rooted ones can simultaneously enhance soil structure, microbial activity and nutrient retention²⁷. Overall, crop effect traits play a central role in shaping soil multifunctionality.

Aesthetic value, heritage and landscape identity

Cultural services emerge from the full range of plant traits that communities perceive, use and value. Variations in morphology, colour, stature and phenology shape landscape aesthetics and contribute to place-based identity⁴⁷. Landraces, often characterised by distinctive phenotypic and sensory traits, are closely associated with local knowledge systems and gastronomic heritage, reflecting long-term coevolution between crops and their socio-ecological contexts^48,49. Maize landraces, for instance, are considered important for their kernel colour, texture and culinary properties linked to traditional food systems^50,51. Maintaining these varieties supports cultural practices, traditional knowledge and place-based values underpinning gastronomic tourism and environmental education⁵². Cultural services thus extend crop multifunctionality into social and identity dimensions that biophysical indicators alone cannot capture.

Assessing crop multifunctionality requires a structured sequence of steps, illustrated with a concrete example in Fig. 2. First, trait–process mapping should be established through literature synthesis, linking plant functional traits to ecological processes and associated ecosystem services. This approach builds on a substantial body of research, largely developed in natural and semi-natural systems, where response–effect trait frameworks have been used to relate plant traits to ecosystem processes^53,54. For example, leaf economic traits and plant height explain variation in productivity and nutrient cycling across plant communities^55,56. Extending such frameworks to cultivated species represents a conceptual advance for crop science^10,39. This knowledge base provides a first filter to identify functionally meaningful traits. Combined with publicly available trait databases, these traits can be used to compare the potential multifunctionality of crop types, species and cropping systems, including annual and perennial crops⁵⁷. Najm et al.⁵⁸ demonstrate this approach by integrating a literature-based synthesis of trait–function–service relationships with large-scale queries of the TRY plant trait database, enabling the systematic evaluation and ranking of service plant species according to their potential contribution to targeted ecosystem services and supporting the design of structured species portfolios for agroecosystem management. Nevertheless, most trait–process relationships are derived from wild species, trait information for crops is often incomplete, and trait–service links can be context-dependent, limiting the reliability of literature-based predictions at finer scales^59,60,61. This highlights the need for field-based validation under real cropping conditions, explicitly accounting for phenotypic plasticity and genotype × environment × management interactions⁵⁹.

Second, field-based tests using a functionally diverse set of crop species or varieties are required to evaluate and refine trait-based expectations. For many crops, particularly underutilised species and landraces, trait information remains scarce, making direct measurement necessary to establish ecological relevance⁵⁹. Once a trait portfolio has been defined, targeted above- and belowground indicators such as pollinator visitation, natural enemy activity, microclimate buffering, soil organic matter, nutrient retention and water infiltration are needed to assess whether trait differences translate into realised service provision. The choice of indicators should align with both production objectives and environmental goals. Such approaches have been applied both in non-crop systems, linking plant traits to pollination, herbivory and biological control⁶², and directly to crops: Dubs et al.⁶³ characterised 16 wheat varieties by 26 traits and quantified 15 services across 88 field plots, demonstrating that specific bundles of variety traits are associated with distinct service profiles. This approach can be simplified for farmers, breeders and advisors by focusing on a small set of locally relevant crop species and varieties, pairing a limited number of easily measurable effect traits with targeted source service indicators. For instance, flowering duration as an indicator of pollinator support, canopy architecture paired with natural enemy abundance as a proxy for biological control, or residue quality traits such as C:N ratio linked to decomposition rates as an indicator of nutrient cycling.

Third, once trait measurements and field-based service indicators have been collected, these data can be analysed to link crop traits to ecosystem service outcomes and measure multifunctionality. Trait–service relationships are first examined for individual services to assess whether specific traits consistently drive ecological processes under field conditions. Multifunctionality is then defined as the simultaneous performance of multiple ecosystem functions and requires integrating several indicators into a common framework⁶⁴. While this concept originates from biodiversity–ecosystem functioning research, it has been increasingly extended to ecosystem services and is here applied to crop multifunctionality. Because functions and services differ in scale, direction and variability, indicators must first be standardised so that higher values consistently reflect higher levels of service provision. Several complementary approaches are available. Averaging approaches combine standardised values into a single index of overall performance, but may mask trade-offs among services^64,65. Threshold approaches quantify the proportion of functions exceeding a predefined level, capturing capacity to sustain multiple services simultaneously, though results depend on threshold choice⁶⁵. More recent integrated metrics jointly capture performance level and evenness by combining mean function value with the effective number of functions, offering greater mathematical rigour⁶⁶. Multivariate approaches use PCA to summarise multiple functions into a composite index that explicitly accounts for inter-function correlations⁶⁷. These approaches can be combined for greater robustness, though none directly quantifies trade-offs or synergies. Service weighting should always be made explicit, as it reflects stakeholder priorities and directly influences multifunctionality estimates⁶⁸. Such methods have been applied across cover crop functional groups¹⁸ and agricultural systems⁶⁹ to identify species and management contexts that sustain multiple services simultaneously.

Fourth, trade-offs and synergies shape the functional balance of crops and must be explicitly assessed^70,71. Trade-offs between yield and regulating or supporting services are a recurring challenge, since services often compete for shared limiting resources, and individual traits can enhance one while constraining another, or jointly support several⁷². A range of analytical approaches can detect and quantify these relationships, with pairwise correlations providing a first assessment of synergies and trade-offs among services across species or varieties⁶³. Multivariate and network approaches identify co-varying service bundles and associated trait combinations, revealing distinct synergies and trade-offs among groups³⁹. Structural equation models can test causal pathways linking traits, environmental drivers and services, distinguishing direct from indirect effects¹¹. Pareto optimality analysis identifies service combinations where improving one service necessarily reduces another, delineating trade-off frontiers across crop species or systems⁷³. Because different methods capture distinct aspects of service relationships, combining approaches provides a more robust understanding of trade-offs and synergies⁷⁴. This can ultimately support the identification of crop species or varieties that minimise trade-offs and achieve favourable balances between yield and ecological service delivery.

Recognising crops as active providers of multiple services makes multifunctionality an operational criterion for crop performance, one that is measurable, comparable across species and varieties, and directly actionable in breeding, management and policy^65,75,76.

Mobilising crop multifunctionality in agricultural practice begins with breeding. Rather than selecting solely for production-related traits, breeding programmes can explicitly target effect traits that enable crops to supply ecosystem services, ideally through participatory approaches that engage farmers and local contexts in which these traits are expressed and valued^32,77. Participatory breeding approaches recognise farmers and traditional knowledge holders as co-creators of agrobiodiversity shaped through long-term human–environment interactions. Trait–process mapping enables the identification of traits underlying specific service contributions and helps anticipate the ecological effects of new germplasm. Genome-wide approaches, supported by pangenomes and DNA variation maps, further offer opportunities to identify alleles associated with ecosystem service provision. In this context, landraces and underutilised crops remain key reservoirs of ecological traits that have been diminished in modern, yield-focused cultivars³⁶.

At cropping-system and landscape scales, multifunctionality can be promoted by selecting and combining crop species and varieties according to specific environmental contexts and management challenges. Besides seeking overall superior genotypes, this approach also recognises that different crops and varieties excel under different conditions and deliver distinct ecological functions. One promising direction is therefore the deliberate design of species and varietal portfolios, including mixtures and polycultures, that combine complementary ecological roles within and across fields. For example, aboveground complementarity can be achieved by combining high-yielding cultivars with nectar-rich landraces to support pollinators, while belowground complementarity can be promoted by pairing deep-rooted genotypes improving soil structure and water infiltration with fibrous-rooted types enhancing nutrient retention and microbial activity. The effectiveness of such combinations often depends on local ecological knowledge, which guides the choice of species and varieties, their spatial arrangement and management timing⁷⁸. Scaling this portfolio-based approach from individual fields to farms and landscapes offers opportunities to enhance overall multifunctionality or to prioritise complementary services across space, particularly under variable climatic conditions and diverse management regimes. Key research priorities include defining context-specific trade-offs, quantifying genotype × environment × service interactions, and developing scalable indicators of service delivery across field and landscape levels. Cultural services represent an additional frontier, as their aesthetic, heritage and identity dimensions require closer integration of ecological and social-science perspectives to be incorporated into crop design and policy⁵².

Realising crop multifunctionality also requires policy and market frameworks that recognise and reward the ecological contributions of crops. Current certification schemes, subsidies and procurement standards rarely account for varieties that support pollination, soil functioning or microclimate regulation, but targeted incentives, such as ecosystem service payments or multifunctionality labels, could promote their adoption. Recognising traditional knowledge and farmer-led stewardship as integral components of agrobiodiversity would further support policies that align ecological performance with cultural and social values.

The perspective developed here leads to a simple but consequential conclusion: crop species and varieties differ substantially in their capacity to provide regulating, supporting and cultural services beyond yield, and this variation is tractable, measurable and largely untapped. The trait–process–service framework makes crop multifunctionality measurable and comparable across species and varieties, transforming it from a conceptual aspiration into an operational criterion for agricultural design. Integrating ecological performance alongside yield in breeding targets, varietal selection and cropping system design offers a concrete pathway to agricultural systems that are productive and ecologically regenerative by design, rather than by accident.